Models: PBA

Portfolio Behavior Analysisadapts CAPM parameters from static "snapshots" into continuous "movies."  A managed portfolio is a changing portfolio.  A changing portfolio is not the same today as it was yesterday.  Changing portfolios require time-responsive analysis.

The rate of portfolio change is often substantial.  Most fund portfolios turn over about once a year, and many turn over much faster.  Conventional betas and alphas (based on 36 monthly returns) may  include performance data from three or more entirely different portfolios.  Successive behaviors are thrown together as if they were one fixed asset.  The result is statistical mush.

PBA addresses this problem with "moving beta."  This construct, originated here in the 1980s, shortens the data span, takes daily instead of monthly returns, and updates parameter values continuously.  Both beta and alpha turn into continuous series instead of a single value...so we get to see a fund's dynamics evolving through time.

Does Beta Variation Matter?  Yes.  It matters because risk matters.  When risk is unstable, your portfolio is unstable.  Two funds may average 1.20, but if one ranges narrowly from 1.10 to 1.30 times market, while the other ranging from  0.4 to 2.0, the fund with wide-range beta is far more risky.  The variation itself presents substantial timing risk.

Does Alpha variation matter?  Yes.  Positive alpha is persistent over weeks or months, but not over years.  Alpha Dynamics captures monthly excess returns by ranking style quadrants for current strength.


Examples

Vanguard Asset Allocation.  Asset Allocation funds are expected to manage market exposures.  VAAPX gives a great example of successful exposure timing.   This chart show the fund's moving beta ("EqExp" in top frame) along with a plot of the S&P to show how the fund's varying exposure interacts with market trends.

VAAPX equity beta (EqExp) varies from about .40 to about 1.00.  Exposure is clearly reduced as the market becomes high-priced, and raised when the market is low.  This pattern shows some anticipation, but is more focused on equilibrating relative value between equity and fixed income.  We estimate this fund adds several percent a year from beta variation.

Magellan Fund  (Link below for 2006 update).  Most growth funds are not considered market "timers."  But when exposure begins to vary widely and sharply, at some point it's clear the fund is making timing bets.  That's what happened with Fidelity Magellan Fund back in the 1990s under Jeffrey Vinik. 

A huge shift into technology cranked net equity exposure all the way to 2.5x market volatility at one point.  Then a radical shift into bonds sent the moving beta plunging...all the way to .40x market.   When the radical mix changes came to light (and weren't playing out well) Jeffrey Vinik's days at Magellan were numbered.

Static metrics didn't capture any of this.  Only investors who were tracking the time series beta (as we were!) could have a clear sense of how much risk and how much timing they were exposed to.

Magellan settled back into a steady Growth Fund style after Bob Stansky took over (and some Fidelity policies were changed) in 1996.  The moving beta exposure settled at about .85 for some months, then finally reverted to minor deviation from general market levels (1.00). 

Update 2006.  Harry Lange took over from Stansky in October 2005.  Portfolio behavior changed dramatically within two months.  Click this Update link to see how PBA revealed the change before the FMR disclosure. 

Moving Alpha Example

Sentinel Small Company.  While beta tracking shows how exposure is being managed, alpha tracking shows how excess returns are being generated.  Alpha is "residual" return (positive or negative) after accounting for the risk-and-market driven portion.  It is generally a measure of effective selection.  Our moving alpha, shows not only how selection's been on average, but how it's varied, and whether it is gaining ground or losing ground today.  Here's a moving "Alfa" picture of Sentinel Small Company Fund (SAGWX)....

SAGWX has generated remarkably consistent positive Alpha over the years.  In the fall of '03, however, Alpha has gone negative while the returns are positive and the NAV is at new highs.  (Prior brief negatives occurred only when the market and fund were down.)  This means although shareholders have still been making money, at Q4/03 they're making it on exposure alone; it's no longer enjoying the "extra" selection kick of strong and growing Alpha.


Style Alphas Moving Alpha also identifies profitable style-shift indications. Top-alpha styles outperform bottom-alpha styles by 2-to-1. For more, click on AlfaDynamics.

For more on the PBA structure and usage (and examples), go to Mutual Funds.  Then for custom subscription to moving beta tracking services, click on BetaStates.

To enquire further about PBA applications, contact us in the space below (or use the multiple checklist back at Models master page)....


To enquire about PBA applications with your own funds and data, you can leave a message back at the Models page, or use the master checklist of Analytix Services.  Or call directly anytime, at (603) 643-6430.

 

Nothing at this site is offered as advice or recommendation.
All site content is subject to full Disclaimer statement and Terms of Use.
Copyright© CarpenterAnalytix.com 2006. All rights reserved.