Mladina on Peeling the Onion: Looking for Idiosyncratic Skill

onionsPeter Mladina, the director of portfolio research for wealth management at Northern Trust, began a recent article with the contention that “diversification is the primary benefit of adding select hedge funds to a portfolio of traditional assets.”

You might respond, “No big whoop.” Hardly radical. But Mladina infers some important corollaries, such as (1) that such investors should be looking for truly diversifying managers and (2) that finding real diversification through the hedge fund world is trickier than it looks. An effort to guide “investors who already own diversified, multi-asset portfolios” through the issue of whether they have the diversification they think they have makes up the gist of his article in The Journal of Portfolio Management (Spring 2015).

Mladina presents his guidance as a “portfolio factor model” of wealth management. The idea – again, one familiar in the literature – is that “only a limited number of compensated risk factors explain the cross-section of diversified portfolio returns empirically. This is where investors should focus.”

One premise is that it can be tough to determine whether you, the investor, receive the benefit of a manager’s idiosyncratic skill versus the benefit of a mix of systematic risk premiums. Ideally, portfolio selection will capture the range of sources of return: from alternative risk premiums, to more exotic risk premiums, to idiosyncratic skill-based returns, “with each perhaps offering different confidence levels relative to traditional risk premiums for forward-looking investors.”

Mladina doesn’t use this image, but it is appropriate to his presentation: Idiosyncratic skill is the core of the onion, a bit that remains after all the skins are off.

Ten Years of Data

Mladina uses the HFRI indices over the 10 year period ending in December 2012 as a database for the table below. Note that in the alpha t-stat row, the event driven category is the stand out. As Mladina puts it, these “low alpha t-stats … show the alphas are indistinguishable from zero for all categories” except for that one.

10 Years to Dec. 2012 Fund Weighted Composite Equity Hedge Event Drive Relative Value Macro
Adjusted R-squared 0.94 0.96 0.89 0.86 0.59
Alpha t-stat 0.81 -1.74 2.19 1.63 1.11
Global Market beta (t) 0.30 0.45 0.28 0.07 0.12
Global Market Beta (t-1) 0.02 0.02 0.05 0.02 0.03
Global Market Beta t-stat 2.13 1.26 2.85 1.39 1.00
Global Credit Beta (t) 0.05 0.03 0.12 0.31 -0.17
Global Credit Beta (t-1) 0.04 0.04 0.05 0.07 0.02
Global Credit Beta t-stat 2.45 2.32 2.26 3.62 0.65

Source: Adapted from P. Mladina, “Illuminating Hedge Fund Returns to Improve Portfolio Construction,” JPM (Spring 2015), Exhibit 7, “Regression Summary with Lagged Market and Credit Factors.”

 

The lagged beta and beta t-stat rows also award event driven the high score, a fact that is “meaningful and persistent” and that suggests to Mladina that “some of the observed alpha may be due to an illiquidity premium the lagged betas do not capture.”

A Nexus of Premiums

Hedge funds do not constitute an asset class, but a nexus of premiums, some of which offer greater value to portfolios than others. Some premiums are available to quite traditional portfolios. Long-only investors can benefit from the global market factor, the emerging-markets premium, the global term, and global credit.  Another five premiums, defined by Mladina as “alternative risk premiums,” are just a bit more exotic: global size; global value/growth; global momentum; commodity momentum; currency premium.  Still others can get much more exotic, as the recent history of financial engineering shows.

Factor models will evolve as researchers untangle what value is to be attributed to what. Model selection, then, has to remain flexible to keep pace with such research, and must of course remain useful for the investment decision makers.

The paper cites work by Sheridan Titman and Cristian Tiu, who found in a 2011 paper published in The Review of Financial Studies that hedge funds that exhibit “lower R-squared with respect to systematic factors have higher Sharpe ratios” and higher alpha. Even more enticing: lower R-squared figures are predictive of future success. The drawback, though, is that these lower R-squared operations are only good bets if the potential investor  is willing and able to do a good bit of due diligence, given the opacity of the sourcing of their returns. Nothing comes from nothing.

Mladina (who holds an MBA from Edinburgh Business School) was formerly the director of research at Waterline, a California based wealth management firm, acquired by Northern Trust in 2010. He is also a guest lecturer on investments at UCLA, and an invited speaker at investment conferences.

 

Be Sociable, Share!

One Comment

Leave A Reply

← Vindication for Pirrong and Irwin: Why Are Trafigura’s Profits Up? Basel: What Does IRRBB Demand from Supervisors or Bank Boards? →