Portfolio Pumping in Singapore — Myth or Reality?

Is portfolio pumping prevalent in Singapore? That’s the question we set out to answer in a first-of-its-kind analysis of this issue in the Asia-Pacific region. We (myself and co-author Tony Tan, DBA, CFA, head of Standards and Advocacy in the AP region) decided to conduct our just-released study in response to the 2010 prosecution of a fund manager involving portfolio pumping.

To gain accurate insight into the degree of portfolio pumping occurring in Singapore, we analyzed more than a decade’s worth of tick-by-tick data from the Singapore Stock Exchange from January 2003 to December 2013. Our study is unique in several ways. It is the first report of its kind in the AP region. (Most of the research in this area has been largely restricted to the western, developed world, until now.) Our research covered 189 companies, spanned 11 years, and analyzed 12 billion data points.

What the Words Told Us

During the literature review process, we uncovered a wealth of research that ranged from merely establishing instances of pumping to identifying favorable conditions for its presence. Key takeaways included:

• The causal relationship between fund manager compensation and the underlying fund performance creates a natural incentive for managers to temporarily boost prices of their fund holdings.
• There are three major theories governing motivation for pumping, as identified by literature. “Leaning for the tape” advocates pumping being driven by the best performers who attempted to maintain their track records despite the general weak persistence of fund outperformance. “Clutching at straws” argues for pumping by the worst performers on fear of going out of business and fund manager anxiety on losing their job. The third theory, “beating the benchmark,” calls for pumping being driven by the marginal players looking to make sure they don’t fall behind the benchmark at the end of the reporting period. There is a wealth of evidence for the first two theories, while the third one has not found many supporters in the academic universe.
• In the absence of the actual details of individual fund manager trades prior to a period-end, most researchers have relied on indirect evidences, drawing from their observation of stock price and fund NAV changes, period-end fund-holding disclosures, and intra-day-trade data provided by exchanges. This is where our tick-by-tick data comes in very handy.
• Academicians have used a continuum of metrics to represent the presence of pumping. At the simplest end, stock or fund NAV returns have been used. Some have observed portfolio disclosures by funds to create implicit pumping measures, while others have used rules of thumb to define a portfolio pumping dummy metric, including checks on returns over the last few minutes prior to a period end and the extent of reversal the next day. Volume-based metrics covering a proportion of block trades to total trades and a fraction of small trades to total trades are among the measures used. Buying intensity and the share of buyer-originated transactions are among the trade-based metrics adopted. At a more complex level, a two-layer regression, involving a regression of residuals in turn obtained by regressing returns against standard style factors, has also been considered in literature.
• The verdict on whether portfolio pumping at the global level is driven by the best- or the worst-performing funds and stocks is mixed, with adequate evidence backing both claims. A broader consensus, however, exists in terms of other characteristics. Pumping is found concentrated among illiquid stocks, predominantly in the small-cap space. Growth funds and stocks have dominated their value counterparts in the manipulation of holdings. Younger funds, less-diversified funds, and small-sized funds with higher turnover, expense ratios, and larger cash reserves have been found as likely candidates. Quite intuitively, evidence suggests that pumping is more prevalent among managers that have underperformed their benchmarks. Team-managed funds fare better than single manager funds. There is also some support for persistence among pumped funds, while persistence is weak at the stock level. In other words, funds that engaged in pumping have a stronger tendency to continue doing so. Each time however, these funds tend to select different stocks to pump.
• Increased regulation, investor awareness, and a robust mechanism to determine closing prices of stocks are among the suggested remedies from literature for policymakers. Last but not least, academicians also call for detailed disclosure of trades, at least of larger trades, especially those executed around a period end.

What Will the Numbers Tell Us?

After our intensive literature review, we proceeded to examine our rich data set to gain further insights. In my next blog post, I will be sharing the key findings and recommendations we found there. Interestingly, the overall outcomes we found are quite different from what we see in the western world.

If you liked this post, consider subscribing to Market Integrity Insights.

Image credit: ©iStockphoto.com/TeamOktopus