More money often does mean more problems for fund managers. For example, a value manager receiving substantial inflows must not only decide how to allocate that extra cash, but must also do so in the knowledge that – because of the sector-led nature of fund flows – plenty of other value strategies will be looking to invest new money into the same types of stocks at the same time.
A new paper indicates that some managers are indeed aware of these kinds of crowding issues and subtly work to minimize the potential problems associated with such behavior.
Vikram Nanda and Kelsey Wei, both of the University of Texas at Dallas, examined the relationships between a fund’s portfolio liquidity, the similarity of its holdings to its competitors in its style category, and the correlation of flows in that fund category. Their sample consisted of actively managed US equity funds from 1993 to 2015.
‘Our tests are motivated by anecdotal evidence in the asset management industry that sophisticated fund managers often watch what their peers invest in and avoid holding the same assets as others, despite the herd mentality,’ Nanda and Wei explained.
Mind the overlap
They constructed a measure they termed ‘overlap management’ to capture dynamic changes in the similarity of a fund’s portfolio to its peers in response to any increased correlation of its flows with those peers.
Broadly, Nanda and Wei found that managers did indeed engage in this type of activity: From the quintile of funds with the lowest flow correlation to the quintile with the highest flow correlation, portfolio similarity dropped monotonically. This trend was highly persistent over time, which suggests that ‘some funds appear to have the skill to manage their portfolio similarity in response to correlated flow shocks.’
By examining funds’ trades, Nanda and Wei discovered that those managers exhibiting the greatest degree of overlap management were less likely to buy or hold stocks that were expected to be subject to flow-driven price pressure when flow correlations in their sector spiked. However, this did not necessarily mean that managers diverted their portfolios into more liquid names when fund flows converged.
‘In fact, funds with greater overlap management tend to end up with portfolios of slightly more illiquid stocks,’ Nanda and Wei wrote. ‘This finding suggests that compared with alternative liquidity-management activities such as increasing the liquidity of holdings and building cash reserves, one advantage of adjusting portfolio similarity is that it could be less costly. In particular, liquidity management through the reduction of portfolio overlap may allow funds to simply substitute overlapped holdings with equally illiquid securities that are less commonly held by peer funds.’
Active share activity
The performance data confirmed this idea. Consistent with plenty of evidence from prior studies of active management, Nanda and Wei found that funds where holdings deviated from the benchmark or their peers do indeed tend to outperform. However, this was not simply about active share: Overlap management contributed independently to higher abnormal returns, with its incremental effect on alpha equivalent to more than two thirds of the effect associated with one measure of active share. That said, the funds that managed overlap dynamically did also display more skill in general stock selection, which may have played a part.
A further advantage of overlap management is that it may help lessen the impact of redemptions on returns. ‘We hypothesize that overlap management could mitigate the feedback effect of outflows on performance and thus the cascading effect of outflows,’ Nanda and Wei said.
‘That is, by avoiding fire sales, outflows would be less likely to result in lower fund performance and subsequently yet more outflows. Supporting our hypothesis, we find that past outflows become less predictive of future fund performance when funds adopt strong overlap management. Correspondingly, we show that overlap management lowers the correlation between past outflows and current flows. Therefore, correlated outflows are less likely to cause a downward spiral in fund performance.’
For fund buyers, one lesson here is that it is worth asking managers whether they monitor their peers’ portfolios. If they do not deign to snoop, they may be doomed to mismanage their own fund.