A managed portfolio is a changing portfolio. As a portfolio changes, its return statistics also change. An evolving asset requires continual tracking to keep current on risk and performance.
Portfolio Behavior Analysis (PBA) tracks performance parameters as continuous time-series. At the heart of PBA are our moving beta and moving alpha constructs. These key parameters have been tracking real-time fund performance since we originated the framework back in 1989. The result is a continuous path of performance dynamics.
Time series tracking shows how returns are generated, not just how much.
Beta measures market sensitivity. But static parameters on a dynamic portfolio don't work right. Most funds have more than 100% annual turnover. So a single-number beta on 36 months can easily reflect three or more entirely different portfolios over the three years. That makes statistical mush.
To better see what's going on, we zoom in on daily returns, and update continuously for a rolling beta series of equity exposure. As return dynamics change, the moving beta changes, giving a continuous picture of risk dynamics.
Example. One fund that varied its market exposure widely and often: T.O. Richardson Sector Rotation Fund.
Popular fund services reported the TRSRX beta at .45, which you can see (above) is about average over the past few years. But "average" isn't worth much when the effective net exposure has varied from ZERO to 1.5x market! Anyone buying this fund expecting .45 beta would be mistaken (except once in a while, by accident).
Most funds don't vary market risk as widely as the Richardson example. But many manage their exposure far more than is supposed. The single-value beta construct fails to meet reality, and fails to show you what you're getting.
Mercifully, this Richardson fund has finally been closed (October 2003), and the sinking NAV path makes clear why.
What is not apparent from the NAV path alone (but is visible in the Equity Exposure line plotted above) is how much of the problem was from bad timing in the form of aggressive and inopportune beta variation.
We do more than track beta variability; we also measure its effect. High beta in rising markets and low beta in falling markets enhances returns. High beta in a falling market can be deadly. (Ask to see a Janus example!) In our moving beta, we can measure these effects, and show you who's doing what!
Timing effect are quantified in BetaGain. This key metric shows who's timing, and who's good at it. In the TRSRX example, the BetaGain model estimates a -5.6% annualized loss from market timing alone. To see a counter example (great timing!) click here on HSGFX-b for a moving beta on Hussman Strategic Growth.
Track your funds' performance and exposures with custom BetaStates profiles. This is not a generic fund service; it's not a newsletter; and you won't see pages of fund talk. It's focused analytical data, and it's customized to your specs.
To enquire or initiate an order -- and to see a sample BetaGain ranking -- click BetaStates.
In a unique extension of BetaGain analysis, TXP shows when fund Tech Sector exposure is shifting to/from better and lesser Tech managers.
Alpha measures stock selection skill. After accounting for market performance, and for a portfolio's sensitivity to market, the residual effect is due to the nature of the specific holdings. That's alpha.
But static single-value alpha (the kind shown in most fund services) is a kind of average value. As with single-value beta, it omits important information; How widely does selection effects (alpha) vary? How much of the time is it positive? Is today's alpha expanding or eroding?
The moving alpha of PBA answers these questions, and allows screening for positive consistency. (That same Hussman fund makes a good alpha example, too; for a view, click HSGFX-a.)
Moving alpha also shows whether a fund's historical capability is continuing. To see an alpha that's decaying into closet index status, click GSECX-a for General Securities Fund.
The same effect of waxing and waning alpha strength underlies the Alpha Dynamics model as it identifies rotations among style funds (with top-alpha styles outperforming bottom-alpha styles by 2-to-1).
(Additional moving alpha and moving beta examples are seen at PBA link in the Models, section.)