ETFs Vs. Mutual Funds: A Case Study - Part III

Includes: IWD
by: Eli Inkrot

In previous articles (found here and here), we have thus far established that the OLVAX has performed very similar to the Russell 1000 Value Index. Moreover, we have found that the benchmark (we use the iShares Russell 1000 Value Index (NYSEARCA:IWD) as a proxy) has provided superior risk-adjusted returns. Let's see if this holds through time.

Manager Returns versus Benchmark graph and table

Manager v Benchmark Annualized
1-year 2-year 3-year 4-year 5-year
OLVAX -0.88% 7.97% 8.29% 9.03% 13.65%
Russell 1000 Value -1.33% 8.23% 9.66% 10.28% 15.07%

The manager returns versus benchmark graph and table (See below) demonstrate how each has performed as of late. As seen in both the table and the graph, in the last year both the OLVAX and the benchmark were negative for the year. It should be noted that this data reflects 2007 data as our last year. I used 2007 because the analysis was from 1997 to 2007. While they were both negative in the last year, the OLVAX was less negative than the benchmark. In the subsequent multi-year periods, for example 2-year, 3-year etc. periods, the benchmark outperformed the OLVAX. It should be noted that these results offer only a snapshot but are consistent with what we observed before, namely the out-performance by the benchmark in the 2003 to 2007 period. If we were to look at the period of 2000 to 2003 the OLVAX would have outperformed the benchmark.

Click to enlarge

Standard Deviation

The standard deviation of the relative holdings can tell the investor about the riskiness of the returns. Here we look at the riskiness through time. (See graph above, below Manager vs. Benchmark returns graph.)

We have already established that the OLVAX has more risk than the benchmark over the 10-year period; however, it can be useful to look at the risk over time. Here we measure risk by using the standard deviation. As you can see, the standard deviation for the benchmark is less than the OLVAX in any year stretch. This is to be expected both because we know that the OLVAX has more risk over time and that the benchmark is likely to be more diversified.

On a risk-adjusted basis over time we find similar results:

Sharpe Ratio

We have already seen that over the 10-year stretch the benchmark had a higher Sharpe ratio than the OLVAX. The first graph below shows a demonstration through time. It is no surprise that the benchmark beats the OLVAX on a risk adjusted basis over time as it is able to do so over entire 10-year stretch.

For those still skeptical, the second graph below tracks the relative performance on a yearly, rather than cumulative, basis.

Calendar Year Return

2001 2002 2003 2004 2005 2006 2007
OLVAX -18.45% -22.63% 34.15% 11.30% 8.92% 17.61% -0.88%
Russell -6.86% -17.18% 36.38% 12.16% 12.58% 18.72% -1.33%

The calendar return table and graph (see above) is truly remarkable in that the benchmark outperformed in 6 of the 7 years. In 2001 and 2002 the benchmark did not lose as much value as the OLVAX. From 2003 to 2006 the benchmark gained more than the OLVAX. As noted previously, the OLVAX did not lose as much in 2007. It should be noted that the OLVAX outperformed the benchmark in 2000. However, looking at the last 7 years of returns, one would see that the benchmark was a much better investment. This is consistent with all of our other findings.


For my case study I used the JP Morgan Large Value fund (OLVAX). I specifically looked into the returns between December 1997 and December 2007. The Fund's objective is to: "Seek capital appreciation with the incidental goal of achieving current income by investing primarily in equity securities". Personally, this phrase right here would turn me off as an investor. My goal was to determine whether or not one could achieve similar results using an ETF versus a mutual fund. Admittedly, this is a singular case and there are obviously are abundance of examples that would suggest that a mutual fund can add value. However, I would stand behind the idea that I did not happen upon this example, insofar as it is inherently different than other potential examples.

As of now, the OLVAX has an annual fee of 1.08%. This compares to the iShares Russell 1000 Value Index, which has an annual expense of 0.20%. In this way, it must be that the OLVAX adds value. From my analysis, this does not appear to be the case.

Looking at the Asset Allocation, both in graph and numerical form, we see that the OLVAX largely stuck to its mandate of large value. Small deviations were found, but this might be explained by definitional or opportunistic discrepancies. With regard to performance attribution, we ran a regression and found that about 94% of the variance in the OLVAX can be explained by the benchmark. This suggests that the manager has very little to do with the overall return in respect to stock selection and weightings.

As seen in the manager performance graph, the OLVAX outperformed in 2000, but underperformed from 2003 to 2007. With regard to the overall return, the OLVAX had a slightly higher return, but took more risk to get there. If you use the Sharpe ratio, you find that the benchmark provided better risk adjusted returns. Also, in comparison to the benchmark, we find that the OLVAX doesn't capture as much upside, but also doesn't capture as much downside. We also used the Sortino ratio which not only looks at the risk-adjusted return, but is also concerned with downside risk. Here, we also find that the benchmark also had better risk-adjusted returns.

The histogram of returns and corresponding distribution statistics are consistent with our other findings and demonstrate that the benchmark performs slightly better. Looking at a cumulative return structure, we see that the benchmark outperformed in the last 2, 3, 4 and 5-year periods. With regard to standard deviation, we see that the OLVAX is more risky than the benchmark in every year stretch. We found no significant alphas for any of the years. The Sharpe ratio over time and Calendar year returns are consistent with our other findings. That is, we see a better risk-adjusted return with the benchmark.

Overall, I would not recommend that someone buy the OLVAX fund. Not only does it largely track the benchmark, but it oft underperforms it. Moreover, the fees are such that one underperforms the benchmark by an even greater amount. Instead, for those looking for a cheap alternative to allocate funds to the large value sector, I would recommend the IWD.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

Additional disclosure: This report, especially with regard to data analysis, was completed with the instrumental help of Kai van Hooff.