Product and Strategy Notes: What is Active Indexing, News for Commodities Asset Class, SEC Investigation and Three New Fixed Income Funds

Information Ratios for Value and Growth Index Tilts

As you might guess, none of these Information Ratios are statistically different from zero (although some of them are close). However, there is an important criticism that can be leveled at this data. There are many different approaches to defining growth and value indices. The ones we have used simply divided the broad market index into two equal halves, based on companies' ratio of book value to market value. Companies with relatively high book/market are in the value index, and those with relatively low book/market are in the growth index. The problem, of course, is that while the difference between companies at the ends of the distribution are quite large, the difference between the companies in the middle are quite small. There is a good argument that this makes indexes constructed in this manner much less useful as a means of realizing what theory suggests should be higher returns from investing in value stocks. A better approach (e.g., the one taken by DowJones Indexes) leaves out many of those companies in the middle of the price/book distribution. The resulting indexes should do a better job of capturing the theoretical value premium. In an upcoming article, we'll analyze the results produced by tilts using these indices, as well as their statistical significance.

By now, the underlying concepts behind "active indexing" should be clear. Based on a theory of the factors that should generate alpha, create a set of rules that enable you to divide the companies in the broad market index into a number of sub-indexes. Then create a product (e.g., a mutual or exchange traded fund) that tracks this index and market it (and the alpha theory) to investors. Over time, you (the sponsor of the fund) earn management fees, and (hopefully) the fund investors earn positive alpha.

In addition to size and value/growth based index funds, investors' interest in active indexing has also led to the creation of index funds that combine these two approaches (e.g., small cap value; large cap growth), index funds that track industry sectors (e.g., consumer staples, financial services, utilities, etc.), as well as even more narrowly defined "indexes" (e.g., the iShares Dow Jones Select Dividend Index Fund, that tracks the 50 shares within the Dow Jones Total Market Index that have the highest dividend yields).

Taken to its logical conclusion, this approach leads to products like the PowerShares Dynamic Market Portfolio (ticker PWC). This fund tracks something called the "Dynamic Market Intellidex Index." In essence, this index is constructed by consistently applying a series of complicated rules that are intended to identify shares that collectively will deliver higher returns that the S&P 500 Index. Conceptually, the only difference between PWC and an actively managed quantitative fund that applies the same approach is the imposition of the intermediate "index." In our view, this is active management, pure and simple, dressed up in index clothing. Has it worked? The fund was introduced in 2003; however, it provides backtesting results covering 1993 - 2003. Over this eleven year period, in comparison with the S&P 500, PWC had positive alpha that averaged 5.145% per year, with a tracking error of 8.90%. Combining these two yields an Information Ratio of 1.92 -- impressive, but still short of statistical significance. More interesting will be the fund's performance going forward. Whether the rules upon which its underlying index is based will continue to generate alpha in the face of evolving changes in the economy and the equity market remains to be seen. And if they do, what is to prevent other smart managers from inferring these rules from PWC's (publicly available) portfolio holdings and trading results, applying them to their own portfolios, and competing away their potential for generating alpha?

As you can see, the basis for a consistently successful "active indexing" strategy is the same as the basis for any other active management strategy: some combination of superior forecasting skill and the number of times you apply it each year. We have repeatedly written about how the former requires either superior information and/or a superior model for making sense of it, and how these advantages are difficult to maintain over the long term. And when it comes to breadth, you face the challenge of balancing the potential additional alpha that you hope to earn from frequent trading against the very real transaction costs that this will incur. In our view, these active management challenges are beyond the skills of most of us, for whom the more prudent course of action is investing in broad asset class indexes.

More Good News About The Commodities Asset Class

A recently published working paper by Gorton and Rouwenhorst ("Facts and Fantasies About Commodity Futures") provides further evidence of why it usually makes sense to have the commodities asset class in your portfolio. The paper analyzes the performance of an equally weighted index of commodity futures that over the period beginning in July, 1959 and ending in March, 2004. This research is relevant to index investors, because the currently available commodities index funds are all based on investments in commodities futures. However, these funds are based on underlying indexes (the Goldman Sachs Commodities Index and the Dow Jones - AIG Commodities Index) that use unequal weights for different commodities (based, for example, on their relative importance in the economy). For simplicity the Gorton and Rouwenhorst study uses an equally weighted index of 34 different commodities. Despite this difference, this study's findings are extremely interesting.
The authors begin by noting that "the economic function of corporate securities such as stocks and bonds is to raise external resources for a firm." In contrast, "commodity futures are quite different; they do not raise resources for firms to invest [in their operations]. Rather, commodity futures allow firms to obtain insurance for the future value of their outputs or inputs. Investors in commodity futures receive compensation for bearing the risk of short-term commodity price fluctuations…Because forseeable trends in spot [i.e., physical] market commodity prices are taken into account when the futures price is set, expected movements in the spot price are not a source of return to an investor in the futures. Rather, purchasers of futures contracts [earn positive returns] when the spot price at the maturity of the futures contract turns out to be higher than expected when they purchased the contract. They lose when the spot price is lower than anticipated. A futures contract is therefore a bet on the future spot price, and by entering into a futures contract an investor assumes the risk of unexpected movements in the future spot price."

However, since "unexpected deviations from the expected future spot price are by definition unpredictable, they should average out to zero over time…What then is the [source of] the return an investor in futures expects to earn? The answer is the risk premium, which is the difference between the current futures price and the [slightly higher] expected future spot price."

The authors then go on to compare the historical returns on their commodities futures index with those on other asset classes. Over the 43 year period they cover, their commodity futures index delivered average annual nominal returns of 11.02% (in U.S. Dollars), with a standard deviation of 12.12%. During this same period, the S&P 500 also had average annual returns of 11.02%, but with a standard deviation of 14.90%. The premium over bond returns for both asset classes was 3.31%.
Equally as important were the correlations of returns the authors found. On ann annual basis, the correlation between stock and commodity returns was negative (.11) -- in other words, when stock returns declined, commodity futures returns tended to rise. Moreover, this negative correlation increased to (.44) when five year holding period returns were compared. In other words, the diversification benefits from holding commodities in a portfolio increase for investors with longer time horizons. To put this in a slightly different context, the authors found that "during the 5% of the months with the worst performance of equity markets, when stocks fell on average by 9.18%, commodity futures experienced a positive return of 1.43%."

The correlation with inflation was also very interesting. For stocks it was negative (.19). While this was better than bonds (.33) correlation, it was still negative -- returns on both asset classes tended to decline when inflation increased. In comparison, the correlation between commodity returns and inflation was positive .31. In short, as a hedge against inflation, commodities were superior to stocks and bonds. Finally, the authors also found that the returns on their commodity index were positively skewed, while those on equities were negatively skewed. The authors concluded that "the slightly higher [standard deviation] of equities, and the opposite skewness [from commodities] together imply that equities have more downside risk relative to commodities."

Just to make sure that their conclusions weren't solely applicable to the United States, the authors repeated their analysis from the perspective of investors located in the UK and Japan. Their results were the same. Finally, the authors tested the performance of their commodity futures index against an index comprising the shares of commodity producing companies. They found that "over the 41 year period between 1962 and 2003, the cumulative performance of the commodity futures index was triple the cumulative performance of the matching equities index." In sum, the authors conclude that "the diversification benefits of commodity futures work well when they are needed most."

Another SEC Investigation About to Explode

We have frequently written in these pages about our frustration with the structure of many defined contribution pension plans. We have three main complaints. First, they usually don't offer a wide enough range of asset classes. Second, they do not offer broad index funds that track those asset classes. Third, too often their default asset allocation usually includes a very high percentage of money market and stable value funds which are inappropriate for most employees. Why do we still find these situations in so many defined contribution pension plans? In the United States, the Securities and Exchange Commission is about to give us an answer. And it may well be a bombshell.

Here's the background. Many pension plans (both defined benefit and defined contribution) hire third party consultants to advise them on critical issues such as asset allocation and the selection of either investment managers (in the case of defined benefit plans) or mutual funds (in the case of defined contribution plans) within each asset class. For this service, the consultants charge the pension plan a fee. In exchange, they have a fiduciary duty to the plan and its participants to provide disinterested advice.

To make a long story short, since last year the SEC has been examining whether the consultants have, in essence, violated this fiduciary duty by taking payments (through various means) from the very investment managers and mutual funds whose services they are expected to objectively evaluate on behalf of their pension clients. While not as obvious as straight-forward kickbacks, these payments have taken the form of practices like charging money managers high fees to attend "educational conferences" sponsored by the pension consulting firm, or selling them advice on how to market their products to pension funds. The problem for the pension industry is that the fees it earns from consulting to pension plans are now apparently dwarfed by the fees it earns from the money managers. In a curious way, this makes sense, doesn't it? Why wouldn't you lowball your bid for a pension consulting contract if your role as a gatekeeper was the key to selling your high priced educational conferences and "marketing consulting" products to a much larger universe of money managers? How different is this from Gillette selling razors cheaply in order to make its real money on the blades? Of course, the consultants might have a leg to stand on if they had fully disclosed these business practices and the potential conflicts they create to their pension fund clients. But few have taken this step.

The potential damage done by these practices could be huge (lawyers, take note). Consider a pension plan with $1 billion in assets. If an inferior manager who was recommended based on its ties to a consultant costs the plan just 1% per year in foregone returns, that amounts to a loss of $10 million each year (not including compounding) for the plan's participants. Like we said, this one looks like it is going to get uglier before it gets better.

Three Interesting New Fixed Income Funds

With relatively few fixed income funds available to investors, the introductions of new ones always catches our attention. Four recent ones are worthy of note. With inflation rising, Salomon Brothers Asset Management (the last firm to carry that once great name, now a subsidiary of Smith Barney, which itself is part of the Citigroup colossus) has introduced a new closed end fund in the United States. Called the Inflation Management Fund (ticker symbol IMF), the fund will hold at least 80% of its assets in real return bonds issues by U.S. and non-U.S. governments, and up to 20% in high yield bonds, including emerging markets bonds. To further boost returns, the fund may employ leverage equal to a maximum of 33% of its assets. Doesn’t this sound like a shrunken version of the old Salomon Brothers proprietary fixed income trading desk made famous in the book Liar’s Poker, by Michael Lewis? And it can all be yours for a front end sales load of 4.5%, along with annual expenses of between .81% and 1.76%, depending on the amount of leverage used. We look forward to tracking its returns.
On the currency front, a firm called Everbank has introduced three certificate of deposit products whose returns are tied to baskets of non-U.S. dollar currencies. The "Viking" is linked to those of Sweden, Denmark, and Norway; the "Petrol" to Norway, the UK, and Mexico, and the "Commodity" to the Australian, New Zealand, Canadian, and South African currencies. Apparently this is what happens when your marketing department trumps your investment department. While we applaud Everbank for trying to make available more products that allow investors to invest in foreign currency fixed income instruments (albeit at the very short end of the yield curve in the case of these CDs), we would much prefer a product that contains a mix of the world’s major currencies, perhaps in proportion to the weights used by the Federal Reserve to calculate the U.S. dollar trade-weighted exchange rate. When it comes to foreign currency bonds as an asset class, we still prefer the T. Rowe Price International Bond Fund.

However, that may change in the future. Pacific Investment Management Company (PIMCO), perhaps the best bond managers in the world, have just introduced a new version of their Foreign Bond Fund. While the existing fund (PFODX) hedges its currency exposure, the new one will not. Apparently PIMCO agrees with the position we’ve long taken, that for investors with long time horizons, hedging this currency risk doesn’t make sense, as it reduces the potential diversification benefits from exchange rate changes.

The final product is the Barclays Advantaged Corporate Bond Fund that has recently been launched in Canada. This is an index product that tracks a weighted mix of 67% of the Goldman Sachs InvesTop Index of liquid, high quality U.S. corporate bonds, and 33% of the Goldman Sachs HyTop Index, which is an equally weighted basket of 50 liquid high yield U.S. corporate bonds. In essence, this is a product for investors who wish to take a credit tilt within the U.S. bond market. What we find particularly interesting is the use of the HyTop Index; we expect it won’t be long before this becomes the basis for a high yield bond ETF in the United States.