In our previous post below—referenced in a story from Jeff Sommer of The New York Times—we performed an analysis of the Laudus Rosenberg US Large Cap mutual fund (sub-advised by AXA Rosenberg) that indicated a significant change in the fund’s risk profile occurred as early as 2008. Our study showed a substantial increase in the daily tracking error of the fund to its benchmark – the Russell 1000 Index. The tracking error reached its peak in June 2009 and was several times higher than many of its quant Large Cap Core peers. So there was a symptom that could have alerted risk managers or investors to potential problems. Note that this symptom became apparent only on shorter-horizon “daily temperature charts” rather than two-year averages commonly used by many practitioners.

We then performed some quick diagnostics to understand the nature of the problem and to eliminate false alarms such as data issues or a specific stock bet. We found none of those but the fund’s daily beta was on a path of steep descent in 2009— another worrying symptom but no diagnosis yet.

To get a better understanding of the nature of the problem we will perform a full multi-factor “metabolic panel” of the fund—daily returns-bases style analysis (RBSA). When performed with the right tools that can filter the daily noise, such an analysis could provide key insights and potential answers to specific questions. Through this daily returns-based analysis, we seek to learn more about attribution of the fund’s gains or losses: what bets were made, when were they made and for how long – and with what impact on both performance and risk? It is important to remember that this analysis does not look at actual holdings. Instead, our analysis seeks to understand a fund’s return behavior by comparing its daily returns to those of various passive indices.

Using daily fund NAVs we performed returns-based style analysis in MPI’s Stylus Pro software of Laudus Rosenberg US Large Cap funds against the S&P 500 sectors. The results of the analysis are presented in the charts below where color bands represent portfolio exposures to S&P sectors and cash (exposure to US Treasury bills).

The change in sector exposure at the end of 2008 is immediately apparent with some significant shifts both before and after. Exposure to Energy, Health Care and Consumer Staples changed in a profound way. The quality of this analysis is very high with R-squared being in the 99% range. Again, please note that the analysis was performed using only fund’s NAV and without use of any holdings information and such a result may not directly imply that the fund indeed had such sector position at the time. The best way to further identify possible bets is to plot the fund’s exposures over the S&P 500 index.

In the chart, the bands above the zero line denote sectors where the fund’s sector exposures are greater than those of the S&P 500 index and the ones lower than zero indicate sector exposures less than the index. In 2009, fund exposure differentials peaked in the 2nd QTR and then started slowly normalizing by the end of the year. Our returns-based analysis shows that the fund may have been significantly overexposed to Health Care (the lowest beta sector) and underexposed to some of the highest beta sectors such as Financials. The increased cash position (green color) may indicate that the fund selected more defensive (i.e. lower beta) stocks within sectors. All of these factors could have contributed to the fund having very low beta behavior in 2009, as identified in our previous post.

Incidentally, this period was characterized by what many referred to as a “dash for trash”. Lower quality stocks that were priced for an Armageddon scenario rebounded strongly when the cycle turned in the second quarter of 2009. Consequently, higher quality names, in general, underperformed these lower quality stocks resulting in the underperformance of many fundamental (both quantitative and non-quantitative) alpha strategies. The impact of this phenomenon on a fund’s performance depended on the extent to which these alpha bets were reined in by risk controls.

And finally, let’s see what RBSA can tell us about fund liquidation. On May 2, the board of directors of Charles Schwab made a decision to liquidate four Laudus Rosenberg funds. The funds were immediately closed to new investors and the liquidation was scheduled for July 30th. The fund is still reporting daily NAVs which make it possible to understand how the liquidation is progressing in terms of the fund’s exposure to major sectors. The chart below, showing daily returns-based exposures of the US Large Cap fund , indicates that the fund is currently behaving as though it is about 65% cash with exposure to just a handful of other sectors.

Having such little exposure to the market over the past several weeks clearly boosted the fund’s performance for remaining investors! Again, we would like to make a disclaimer that this analysis is based only on the fund’s NAVs and may not reflect the fund’s actual positions.

The question many investors are now asking is: Does BRK.B actually belong in the Financial sector? If BRK.B behaves like a Financial, for all practical purposes (including portfolio construction and risk management) it should be treated as a Financial. However, our analysis of BRK.B’s historical returns shows it behaves more like a Consumer Staple stock than a Financial.

BRK.B’s sector classification is of particular importance to both long-only and long-short portfolio managers attempting to manage sector exposure. For example, consider a long-short portfolio manager attempting to construct a sector neutral portfolio. If the portfolio manager is long BRK.B and it behaves like a Consumer Staple and not a Financial, this portfolio will have unintended, potentially large sector tilts towards Consumer Staples and away from Financials.

Based on monthly return correlations, it appears that, in recent years, BRK.B behaves more like a Consumer Staple than a Financial.

As the chart below clearly shows, not only are BRK.B’s returns more closely correlated with Consumer Staples, but the magnitudes of its returns are more similar to Consumer Staples than Financials.

The results from this simple correlation analysis were further supported by more rigorous, quantitative analysis done using MPI’s proprietary Dynamic Style Analysis (DSA) model. When using a multi-factor model comprising of the S&P 500 sectors, Consumer Staples is the dominating factor driving BRK.B’s returns. Since 2001, the exposure of BRK.B’s returns to Consumer Staples is much larger than its exposure to Industrials and Financials combined.

Why does the market treat BRK.B as a Consumer Staple instead of a Financial when, according to Berkshire itself, their primary business is in the insurance and reinsurance sector? We posit that three different considerations, when viewed together, provide a plausible explanation. First, a large chunk of Berkshire’s insurance businesses are viewed as Consumer Staples not Financials. Second, Berkshire’s reinsurance businesses impact BRK.B’s idiosyncratic returns, but do not have a significant impact on its systematic returns. Finally, and most obviously, Berkshire has large public holdings in companies like Coca-Cola and P&G.

Interestingly, looking at the exposure chart above, prior to 2001, BRK.B traded more like a Financial than a Consumer Staple. What explains this apparent shift in how the market perceives Berkshire Hathaway? We will continue our investigation into BRK.B and Berkshire Hathaway in future blog posts.

Using “traces in the sand” we will try to recreate the pieces of the performance puzzle that are otherwise hidden from the investors. These traces represent the fund’s monthly performance numbers, which frequently the most the investors would get from a fund. The chart below represents a dynamic analysis of the fund using Russell style indices and MSCI EAFE as a proxy for Int’l equities.

Although the exposures remain similar to the ones in our 2007 study, one could observe a profound change: the fund is behaving as if it’s net neutral or net short. This could be inferred from the sum of short exposures (about 90% in midcap growth) being about the same or greater than the sum of long exposures above the X-axis. That’s quite a change as compared to exposures 3-4 years ago which indicated a 100% net long allocation. No wonder the fund missed the recent market rally.

Moreover, in April alone, midcap growth stocks (short exposure in REIF) delivered twice the performance of large cap stocks (long exposure) which is reminiscent of the statement about the volatility in the fund conference call. Basically, our analysis of the fund’s returns shows that capitalization is the most dominant factor of the fund’s performance. The chart below translates it in the language understood by finance professionals: performance attribution. Based on our analysis, RIEF April losses are dominated by its net short exposure to midcap growth stocks.

As a reference we provide the following chart that shows how well the portfolio based on exposures “Style” is replicating (in-sample) the fund (“Total”).

Please note, at no time in this analysis are we claiming to know or insinuate what the actual strategy, positions or holdings of this fund were; nor are we commenting on the quality or merits of Renaissance’s strategy or that of any other manager. Instead, we are seeking to demonstrate how MPI’s Dynamic Style Analysis can be used to better understand fund behavior, anticipate performance, identify risks and, possibly, replicate fund performance in certain cases.

This neither legitimizes SIB nor makes the stable value fund an immediate suspect. It is just another confirmation that investors need more disclosure on investment vehicles priced at book value. Having credit and interest rate exposure information is important. Having monthly market values would be extremely helpful. Not only it allows one to monitor the the MV/BV ratio, but more importantly, helps to uncover hidden risks using returns-based tools.

There is growing public awareness of potential risks associated with book value accounting in stable value funds. A couple of weeks ago I mentioned my blog to Eleanor Laise with The Wall Street Journal and she wrote an excellent article on stable value funds in the March 26 issue of WSJ.

1. Please see my previous blog post

In the February 25, 2009 SEC complaint, Paul Greenwood and Stephen Walsh were charged with defrauding investors of $554 million through WG Trading Hedge Fund (“WG”). WG employed an index arbitrage strategy and charged 1% management fees and 20% performance fees. When we evaluated WG’s monthly return series’ statistical properties, we found them to be very similar to the average money market fund.

Over the past 5- and 10-years WG produced annual returns of only 50bp higher than some money market funds. In addition, the correlation between WG and money market funds is around 90% and for some funds is as high as 98%! Lesson learned: a simple statistical profile, in this case correlation analysis, can provide a useful sanity check before committing millions of dollars to a particular investment strategy.

Robert Stanford’s SIB marketed stable high-yielding CDs to investors as a part of his alleged fraud. Returns were in the 6-10% range¹ and were significantly higher than CDs offered by commercial banks but very similar to the returns on the investment product that many of us have in our respective 401(k) and IRAs – Stable Value funds (SVF).

Using data provided in the SEC complaint we compared statistical properties of SIB returns paid to investors with a large group of stable value products and found them to be almost identical (based on 15 annual return numbers) both in risk and return. Moreover, the correlation of some of these funds to Stanford was as high as 94%.

In all recent cases of alleged fraud including Madoff, Stanford and WG Trading, investors were looking for stable positive returns and were willing to take risks and invest in investment products with no protection and no transparency. Stable Value funds, while providing very similar smooth positive returns, are legitimate registered products² but investors have to understand that they carry certain risks. Unlike money market funds, SVF invest in longer-term high grade bonds and, therefore, are exposed to interest rate and credit risk. At the same time, these risks are not visible to investors because SVF are priced at book value rather than market value which allows them to report such stable positive returns. Principal and accrued interest are guaranteed by a number of insurance-wrapper contracts with multiple insurers AIG included. As of today, almost 20% of all DC assets or over $400B is invested in SVF.

There are several links below that go into more detail on the risks of SVF and the call for more fiduciary rigor in their assessment.

David Merkel’s Blog
A very good insight into structure and risks of stable value products from an industry insider where he’s pointing to interest-rate, credit and asset default risks.

Stable Value Investment Association
Describes challenges facing these funds in the current market environment. In short, investors should request better transparency, especially on MV/BV ratio and wrapper agreement terms.

J.P.Morgan Insight > Retirement Plans
Examines the procedures plan fiduciaries use in evaluating their stable value fund options. Has a fund performed well if the stabilized returns are positive and stable? You can only assume. No plan sponsor can clearly determine what risks were taken and whether the manager’s returns reflect a prudent tradeoff of risks and rewards by simply looking at the book value returns.

1. Over the period 1992-2006. See chart on p.28 of the complaint
2. With an important caveat: stable value funds are registered with DOL, not SEC.

In the January 3^rd 2009 article “Mutual Fund Fought Off Bears but Now Is Clawed” The Wall Street Journal reporter Eleanor Laise attempts to find an explanation of the Fairholme fund’s subpar performance in the last quarter of 2008. She interviews fund manager Bruce Berkowitz as well as industry experts, analyses fund holdings, and looks at the most recent stock picks. Her conclusion: the fund made all the right sector moves (e.g., got rid of energy and financials) and its otherwise exceptional performance in 2008 has been marred by just a series of poor stock bets. The article concludes with assurances from the fund manager that the strategy is sound and that all the bets will pay off. We also learn that Mr. Berkowitz now has almost 100% of his net worth invested with his fund. The latter is indeed a good sign but we believe that a sophisticated investor deserves better insight on what the future might hold for this outstanding fund.

By performing an analysis of the Fairholme return record using MPI Stylus we tried to connect the dots to understand the sources of the fund’s past performance and provide some guidance in what to expect in the nearest future.

“Ignore the Crowd”

Fairholme fund indeed has an exceptional track record. It outperformed the S&P 500 index every single year since its launch at the end of 1999 except for 2003. Even last year, which was the focus of the WSJ article, the fund outperformed the index by 7.3%.

Fairholme is a no-load mutual fund with an expense ratio of 1%. Its manager is true to its motto “Ignore the Crowd” as prominently featured on the fund’s website. The fund makes focused sector and securities bets. It’s concentrated, holding about 22 stocks. The top ten holdings account for 64% of its assets (according to August 31, 2008 data.) We consider this degree of concentration to be especially high given the fund’s size of $7.2B. Annual turnover is only 14%.

Fairholme is classified as Large Blend by Morningstar and Mid-Cap Growth by Lipper. This seems a little odd, though it’s not surprising for fund research firms to disagree on such a concentrated portfolio. We performed a returns-based style analysis of the fund to better understand its style exposure. Then we undertook a second analysis with its focus on sector exposure to get at its underlying strategy and to determine the sources of its past and recent successes. This put us in a position to either confirm or refute conclusions presented in the reporter’s story.

In the chart below we show the fund’s cumulative performance since inception vs. the S&P 500 Index. A chart similar to this is prominently positioned in Fairholme’s semi-annual 2008 and annual 2007 reports. Note the accentuated 2008 loss—that’s how performance appeared to long-term Fariholme investors.

The depiction of the drop in 2008 seems a bit confusing given that Fairholme outperformed the index by 7.3% that year. For this very reason we find such “growth” charts to be very misleading because the compounding of a single-period return outlier gets multiplied and creates a distorted view of a fund’s performance. A simple cure for this problem is to employ a logarithmic scale on the y-axis as in the chart below. Doing so shows that most of the gains are attributed to years 2000-2002 while over the past several years the fund’s performance was more or less in sync with the S&P 500 index.

Fairholme could help itself either by replacing their growth chart in the 2008 annual report with this alternative form or by showing the fund’s annualized performance next to it. Such a chart will give the fund deserved credit for its outstanding track record despite recent losses.

Getting Under the Hood

In order to determine the drivers of the fund’s performance we needed to perform a returns-based style analysis (RBSA) of the fund. As a reminder, RBSA attempts to create a dynamic portfolio of generic asset indices that replicates the analyzed fund’s performance. For our analysis we use the nine years of Fairholme monthly returns¹ through the end of 2008, the six Russell style indices, and the MSCI Canada equity index to represent Fairholme’s significant exposure to Canadian stocks. The results of the analysis are presented in chart below.² Note that we didn’t use any holdings information to produce these results. All we needed was a stream of monthly return figures.

The fund behaved as if it held a significant amount of cash that it unloaded entirely in recent months. This could mean that the fund indeed held treasuries, but it is also possible that it included some financial or other interest rate sensitive stocks. SEC filings by Fairholme confirmed that the fund held 25% or more in treasuries, money market funds and bonds during this period. The rest of the cash exposure could be explained by the relative defensiveness or interest rate sensitivity of the specific stocks in the Fairholme portfolio vs. the generic style indices. In any event, this definitely contributed to the fund’s outstanding performance during 2000-2002. Note also defensive style of the fund – its lack of exposure to growth stocks. An exposure to Canadian equities built up during 2005-2007 diminished significantly by the end of 2008. No doubt long-time investors in the fund were expecting the fund manager to make a bold move out of the market to weather the financial storm the way it did back in 2000-2002. But let’s not forget that the size of the fund at that time was a mere $50M vs. $7B+ today!

The quality of the analysis is good with the R-squared equal to 79%, which is relatively high for such concentrated portfolio. It is worth noting that RBSA is especially valuable for studying concentrated portfolios such as Fairholme whereas an analysis based solely on holdings could sometimes produce a distorted view. For instance, an industry sector could be represented by a single stock, the performance of which may have little relationship to its sector’s performance, so provide misleading information about sector representation to an investor.³ Rather than using accounting information and what is sometimes a subjective industry classification, RBSA looks to the effect of a stock or, more properly, all of the stocks on portfolio behavior as viewed and valued by a consensus of market participants.

Another reason why RBSA is particularly helpful in the analysis of concentrated portfolios is that managers of funds with few holdings sometimes engage in “window dressing” to present their portfolio in an appealing way right before the reporting time. An analysis of the return record of a fund uncovers the real drivers (exposures) of its performance, so is not affected by such holdings manipulation.

From the Style Map below we can conclude that Fairholme’s US equity position has been generating performance consistent with Russell Midcap Value Index and that its capitalization size has decreased in the recent years. This is quite different from both Morningstar’s and Lipper’s classifications placing the fund in Large Cap Blend and Mid Cap Growth categories, respectively, based on its portfolio information. As mentioned above, our mapping is not derived from holdings but is based on the fund’s performance relative to Russell style indices and depends on the Russell style classification, so we could expect some differences between both. But this difference between holdings-based and returns-based style is quite significant and should be noted as it puts us in a better position to assess the fund’s near-term risks and exposures.

Daily Analysis

Though an analysis based on monthly frequency return data is sufficient for a reasonable assessment of the fund’s long-term results, we need to resort to the use of daily frequency return data to get a better understanding of the fund’s recent behavior. By having more data points to enrich the same or a shorter return history, we are able to use more indices (explanatory variables), such as economic sectors and sub-sectors, to get a more detailed analysis. For the daily frequency analysis of the fund’s performance in the 4^th quarter of 2008, we used Fairholme’s daily return record as provided by Lipper/Reuters, along with daily returns of MSCI Canada and the S&P 1500 GICS sector indices for the same time period. Given the results of our monthly analysis, we thought that a broad equity index such as the S&P 1500 Index would better represent the fund’s positions than the large cap S&P 500. The results of daily RBSA covering the last quarter of 2008 are presented in chart below.

Although our results do not aim to represent the fund’s holdings, they could be defined as “implied holdings”, i.e., a sector combination that best explains the fund recent performance. In many cases such implied information agrees with the holdings report, especially for highly diversified funds holding hundreds of stocks. Here, the diminishing cash and foreign exposure is consistent both with the WSJ article and the monthly frequency analysis presented above. Negligible Energy and Financials exposures are also consistent with the reported data. One immediate observation from the chart is that the fund manager is making fairly concentrated sector bets and the fund’s performance is dominated by the two sectors: Health Care and Consumer Discretionary. While the significant Healthcare component was noted in the article, the sizable exposure to the Consumer Discretionary segment is rather surprising. Companies in this segment tend to be much riskier and more susceptible to market downturns than in the Consumer Staples segment. This could have contributed to the fund’s poor results in the 4^th quarter. The differences in sector exposures vs. the market index are better seen in the chart below where positive values indicate over-exposure while negative ones (below X-axis) suggest under-exposure. Note that the fund seems to be consistently over-exposed to the two above mentioned sectors and is under-exposed to the rest of the sectors.

Fairholme underperformed the S&P 1500 broad equity index –by 1.64% in the last quarter of 2008. Instead of trying to guess which of the sectors affected the fund’s results in the last quarter of 2008, we performed an attribution analysis to quantify the impact of each bet. We present results of this analysis in the “Excess Attribution” chart, which represents a breakdown of the 1.64% loss vs. the index. Each bar in this chart shows the impact of over/under weighting a sector in the portfolio vs. an “index-neutral” bet, i.e., passively investing in the index proportional to its weight within a broader index.

In a quarter when all sectors experienced losses, the most significant positive contribution came from the fund’s small cash position. Other notable positive bets were the overweighting of Health Care stocks and being out of Financials. The most important observation is that the “Selection” component accounted for a loss in the quarter of well over 5%. Selection return represents the difference between the fund’s performance and the performance of the dynamic portfolio of sector indices (shown in the Asset Loadings chart above) and is usually attributed to the manager’s within-sector security selection ability. In this particular quarter the portfolio suffered the most from specific stock picks within sectors rather than from exposure to any particular sector. So, in this regard our findings echo the WSJ article, which targeted specific stock underperformance.

In summary, fund investors should take away from our analysis the following: (a) sector bets overall proved positive during the quarter (although they should keep in mind the exposure to Consumer Discretionary stocks as we’re moving deeper into a recession) and (b) negative Selection bets in a single quarter may prove positive down the road as the fund manager is usually making long-term bets.

1. Fund returns are provided by Morningstar
2. The analysis is performed using MPI Stylus software in DSA mode. For interpretation of this and following charts please refer to the “Style Primer” and other white papers in the Research section of MPI’s website.
3. It’s a known fact that there are stocks in each economic sector that have higher correlation with sectors other than the ones they belong to.