Morningstar Advisor - February/March 2009 - (Page 36) Spotlight Exhibit 2: Turbulence Year-End 2007 Std Dev 5 Yr Std Dev 10 Yr Kurtosis 5 Yr Kurtosis 10 Yr MS Risk 5 Yr MS Risk 10 Yr earlier. Dodge & Cox Stock even had a negative five-year kurtosis value. Investors were offered no indication of the volatility that was beginning to occur in the funds’ portfolios. Elevated Risk SSga Yield Plus Fidelity Ultra-Short Bond Ultrashort-Bond Category Avg Year-End 2007 3.85 2.21 0.76 2.85 * 0.80 Std Dev 5 Yr 19.83 12.31 1.30 40.47 * 1.02 Kurtosis 5 Yr 0.16 0.05 0.01 0.09 * 0.00 MS Risk 5 Yr Regions Morgan Keegan Select High Income High-Yield Bond Category Avg Year-End 2007 Std Dev 5 Yr 17.55 4.66 Std Dev 10 Yr Kurtosis 5 Yr 13.19 1.41 Kurtosis 10 Yr MS Risk 5 Yr 3.84 0.24 MS Risk 10 Yr Dodge & Cox Stock Large-Value Category Avg Year-End 2007 9.16 8.89 Std Dev 5 Yr 13.83 13.28 Std Dev 10 Yr –0.27 0.35 Kurtosis 5 Yr 1.61 1.56 Kurtosis 10 Yr 0.91 0.84 MS Risk 5 Yr 2.06 1.84 MS Risk 10 Yr By month-end November 2008, the markets and the economy were in a full-blown crisis, which is evident in the risk data in Exhibit 3. Once again, we see that the Morningstar measure registered the sharpest risk penalties for the bond funds. Likewise, the risk measures of both Dodge & Cox funds took a dramatic turn as the funds’ troubled financials holdings were forced to sell or seek government bailouts. Hard Lessons Dodge & Cox Balanced Mod Allocation Category Avg 6.23 5.92 8.86 8.91 0.23 –0.04 1.29 1.04 0.41 0.37 0.82 0.82 *Fidelity Ultra-Short Bond did not have a 10-year record at the end of 2007. Exhibit 3: In Free-fall November 2008 Std Dev 5 Yr Kurtosis 5 Yr MS Risk 5 Yr Our analysis shows that portfolios that contain thinly traded or complex derivative securities— or funds that own companies that are exposed to these types of investments—have risks that can escape detection for years. If investors rely solely on quantitative risk measures such as the ones discussed here, real risks can escape detection until the damage has been done. True, measures such as kurtosis and Morningstar Risk are more sensitive to damaging outliers than standard deviation. Still, the key drawback for all these measures is that they are backward-looking. The complex derivative securities owned by the three bond funds, for example, did not have enough history for the risk measures to pick up their dangers, leaving investors in the dark about the true downside potential of these funds. The bottom line is that it’s important to remember that such measures can’t replace the value of understanding what risks might be harbored that could materialize, but haven’t yet. Even so, with financial engineering taking a breather, one hopes that the science of risk measurement will get a chance to catch up. K Arijit Dutta is an associate director of mutual fund analysis with Morningstar. Fidelity Ultra-Short Bond Ultrashort Bond Category Avg November 2008 3.08 2.22 Std Dev 5 Yr 8.51 10.99 Kurtosis 5 Yr 0.09 0.07 MS Risk 5 Yr Regions Morgan Keegan Select High Income High-Yield Bond Category Avg November 2008 Std Dev 5 Yr 24.96 10.24 Std Dev 10 Yr Kurtosis 5 Yr 2.06 15.32 Kurtosis 10 Yr MS Risk 5 Yr 4.99 1.31 MS Risk 10 Yr Dodge & Cox Stock Large-Value Category Avg November 2008 15.17 13.18 Std Dev 5 Yr 15.62 14.14 Std Dev 10 Yr 5.47 6.21 Kurtosis 5 Yr 2.79 2.32 Kurtosis 10 Yr 2.46 1.82 MS Risk 5 Yr 2.58 2.02 MS Risk 10 Yr Dodge & Cox Balanced Mod Allocation Category Avg 11.52 9.78 10.99 9.90 7.22 8.03 4.98 3.36 1.39 0.99 1.28 0.99 weighs positive and negative deviations equally (kurtosis does, too), whereas Morningstar Risk penalizes only downside variation, and heavily so. Thus, standard deviation, the most frequently used risk measure, is likely to understate the true downside of a portfolio. Little had changed, riskwise, for the Dodge & Cox funds at year-end 2007. The crisis had yet to touch the funds’ financial holdings, which had little direct connection to the mortgage blowup. In fact, both funds and their category peers looked less risky than they did a year 36 Morningstar Advisor February/March 2009
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.