Today’s column in MarketWatch describes the downfall of the Sequoia Fund (SEQUX). The fund sports an exceptional long-term record:

Over 45 years through the end of 2015, Sequoia Fund has returned 14.0% per annum versus 10.8% for the S&P 500 index, with $10,000 invested at Sequoia’s inception worth $3.9 million, versus $1.0 million for the same amount invested in the index. Over the last twenty years, we have returned 10.2% per annum, versus 8.2% per annum for the index.

However, over the past eight months this non-diversified fund was undone by an excessive position in just one stock, Valeant Pharmaceuticals International Inc. (VRX). How did the fund fare overall from the perspective of Alpholio™’s patented methodology?

For each analyzed fund, Alpholio™ constructs a reference portfolio of ETFs that mostly closely tracks periodic returns of the fund, thus adjusting for the fund’s volatility (to learn more, please visit our FAQ). Let’s start with the simplest variant of the methodology, in which both ETF membership and weights in the reference portfolio are fixed over the evaluation period. Here is a chart and related statistics of the cumulative RealAlpha™ for the Sequoia Fund in this analysis mode:

Cumulative RealAlpha™ for Sequoia Fund (SEQUX) - Regular Fit

From late 2004 through 2010, the fund generated hardly any RealAlpha™ over its fixed reference ETF portfolio. While the cumulative RealAlpha™ peaked at almost 82% in July 2015, it has since declined to less than 21% in February 2016. Consequently, the fund produced only about 1.5% of annualized discounted RealAlpha™ over the entire period. The standard deviation (a measure of volatility of returns) of the fund was significantly higher than that of the reference portfolio; this is consistent with the concentrated nature of the fund’s holdings.

In a more elaborate variant of Alpholio™’s methodology, the membership of the reference portfolio is fixed but the weights can fluctuate to more accurately track the fund’s returns over time. The following chart and statistics illustrate this outcome:

Cumulative RealAlpha™ for Sequoia Fund (SEQUX) - Fine Fit

In this analysis mode, the fund added only about 1% of annualized discounted RealAlpha™. The standard deviation of the reference portfolio increased to about 12.5%, closer to 13.7% of the fund. (Note that in this analysis, the first data point was in February 2005, i.e. three months after the start of input data. This lag stems from the sliding time-window aspect of Alpholio™’’s methodology.)

Finally, in the most advanced variant of Alpholio™’s methodology, both the ETF membership and weights in the reference portfolio may change over the analysis period. Here is the resulting chart with accompanying statistics:

Cumulative RealAlpha™ for Sequoia Fund (SEQUX) - Detailed Fit

Compared to this reference portfolio, the fund produced a negative annualized discounted RealAlpha™. This means that many of the substantial bets the fund made on its holdings did not really add value when fully adjusted for risk. After the fund’s cumulative RealAlpha™ peaked in September 2015 at about 4%, it subsequently slumped to below minus 21% in February 2016. The lag cumulative RealAlpha™, which simulates a substitution of the fund with a reference ETF portfolio with a one-month lag, shows that at the end of the entire evaluation period the fund added only 3% of value through security selection (as of this writing, the fund returned an additional negative 7.5% in March 2016). It is also worth noting that in this analysis mode, the top equivalent position of the fund, in Market Vectors Pharmaceutical ETF (PPH), peaked at at a weight of almost 58%. This underscored the non-diversified nature of the fund with its inordinate exposure to just one industry.

The prospectus benchmark for Sequoia Fund is the S&P 500® Index. One of the long-lived and low-cost implementations of this index is the SPDR® S&P 500® ETF (SPY). Alpholio™’s calculations show that from December 2004 through February 2016, the fund returned more than the ETF in approximately 86% of all rolling 36-month periods, with a median cumulative (i.e. not annualized) outperformance of 34.6%. Similarly, the fund returned more than the ETF in 69% of all rolling 24-month and 60% of 12-month periods. However, as the above analysis clearly demonstrated, merely comparing fund and single-benchmark returns paints a vastly incomplete picture of the fund’s performance.

To learn more about the Sequoia and other mutual funds, please register on our website.

Pin It