Alpha Myth Unicorn

Top Performing Fund Manager For Last 40 Years: You Are Using the Wrong Benchmark

Alpha Myth Unicorn

Imagine after 40 years of being a professional money manager, you look back and say, “I was the chosen one.” Making it through events such as the Oil Crisis of 1973 and the stagflation of the 1970s, to the falling of the Soviet Union, globalization, and Black Friday of the 1980s, to the booming 1990s and the failure of Long Term Capital Management, to 9/11, the wars in Afghanistan and Iraq, and most notably the last financial crisis and Great Recession that followed; the events which have defined our society in the 21st century. Through all of it, you came out on top. It’s a great story.

A recent article in Financial Advisor Magazine tells the story of Albert Nicholas, the principal portfolio manager at the helm of The Nicholas Fund (NICSX). Mr. Nicholas began the fund back in 1969 and, as FA Mag puts it, “topped the Standard & Poor’s 500 Index by an average of 2 percentage points a year for the past 40 years and has beaten it every year since 2008.” Extremely impressive, indeed!

The article goes on to snub academics who have for a very long time suggested that most active investors will fail to accomplish such a feat. It even suggests that this should be taken as evidence that there is still hope for those who are seeking “alpha” for their investors. The efficient market hypothesis is an idea for cavemen.

It is articles like this that inspire us to wake up everyday and do what we do, to educate and maybe even protect.

Just An Illusion

While it is very common to compare investment results to the S&P 500, it is extremely misleading. This is what we would call the “pulling a rabbit out of the hat” in the active investment community. It is called improper benchmarking. Taking your portfolio, which is made up of many stocks that lie outside the S&P 500 and then compare it to the S&P 500. Why would you do that? To enhance your track record of course.

Let’s start with what we know. We know that although stocks tend to go up and down over time, they do not all go up and down at the same time or by the same amount. Different groups of stocks, better known as asset classes, can be categorized by certain traits such as their size or their relative price (better known as growth or value). Looking over the last 88 years, small cap stocks have been shown to outperform large cap stocks across all different markets around the globe and across different time periods. Similarly, value stocks have outperformed growth stocks across different markets and time periods. With that said, we should expect that any manager who has more exposure to companies that are smaller in size or more value oriented are expected to outperform the S&P 500.

Anyone who has been in the portfolio management business long enough or who has taken professional certifications like the CFA® (Chartered Financial Analyst) or a graduate degree program like an MBA in Finance has heard of the Fama/French 3 Factor Model and are well aware of these premiums. So if active managers are seeking smaller stocks and value stocks, then making the comparison to the S&P 500 is by no accident. 

The Nicholas Fund describes its investment objective as striving to, “increase the value of your investment over the long-term by buying a diversified portfolio that includes primarily medium-and-large size companies. Small companies may be considered.” Also, “our goal is to find growth companies that are reasonably priced by identifying stocks that have low price-to-earnings ratios relative to their earnings growth.” In other words, they have indeed looked into smaller companies than those found in the S&P 500 and they use the PEG Ratio to find value. The PEG Ratio takes the Price-to-Earnings Ratio of a particular stock and divides it by the stock's annual earnings-per-share growth. All of these elements are easily found in companies’ financial statements, so Nicholas must believe that the market has not already priced in these elements even though financial statements for publicly traded companies are, not surprisingly, public. Either way, the Nicholas Fund is using different financial ratios to find value, which may increase its exposure to value stocks at any given time. 

The Analysis

Given our general curiosity for this type of analysis, we took a deep dive into the historical performance, asset class exposure, and historical sensitivities of the small cap and value premiums to see if the Nicholas Fund’s story is what the active investment community is touting it to be. Let’s first start with comparing its performance against its Morningstar assigned benchmark, which happens to be the Russell 1000 Growth Index (a large cap growth index).

Below shows the annual returns of the Nicholas Fund versus the Russell 1000 Index since 1979, the very beginning of the Russell 1000 Index.

As you can see, there are certain periods when the Nicholas Fund beat the Russell 1000 Index and vice versa. From 1979-2015, the average alpha has been 0.79%. Because there is significant volatility in the alpha over time, we want to calculate the t-statistic (measure of statistical significance) to ensure that we are not being fooled by randomness. Given the 37-year history, an average alpha of 0.79%, and a standard deviation of alpha at 10.29%, the t-statistic is 0.47. What this means is that we can be less than 20% confident that this performance is in fact a display of skill and not just random luck. Ideally, we would like to see somewhere in the vicinity of 95-99% confident when advising clients with their money which correlates to a t-statistic of 2.0 and 2.60, respectively. In order to reach this level of confidence with an average alpha of 0.79% and a standard deviation of alpha of 10.29%, we need approximately 683 years of data in order to be confident at the 95% confidence level (still a 5% chance that our conclusion is wrong). We may inhabit another planet by the time this happens.

The Big Let Down

As these results indicate, once we have a better benchmark, the performance is not as impressive as it was originally stated. But this type of basic analysis is not robust. More often than not, most active managers change their overall allocation overtime depending on where they think they can find the most value. It is very unrealistic to assume that the Nicholas Fund has been focusing on large cap growth stocks since its inception.

The good news is that we have the ability to see the historical style drift for the Nicholas Fund overtime through multiple regressions. The chart below gives a visual representation of the general asset class exposure of the Nicholas Fund since 1982 using the Russell 1000 Growth, Russell 1000 Value, Russell 2000 Growth, and Russell 2000 Value indices.

As you can see, in the middle of 1982, most of the exposure was to small cap value stocks. By the early 1990s it had switched to large cap growth stocks, then large cap value stocks, back to large cap growth stocks, and now is mainly comprised of large cap value and large cap growth stocks. This is important. We can see that comparing their performance against the Russell 1000 Growth Index in 1982 would be bad benchmarking. It would have been more consistent to compare its performance against the Russell 2000 Value Index. This is of no fault of Morningstar. They look at current fundamental ratios such as price-to-earnings and price-to-book as well as median and average market capitalization within the fund in order to assign a benchmark. Fortunately, we can go back to our statistical tool bag to find a potential solution.

Proper Benchmarking

The last part of our analysis includes using a multiple regression analysis to find the fund’s average exposure to the size and relative price premiums made popular by the Fama/French 3-Factor Model, which was introduced in 1992. Here is what we found.

Since inception, the fund had a positive loading on both the Small-Minus-Big Factor (small cap premium) and the High-Minus-Low Factor (relative price premium) and both loadings are statistically significant at the 95% confidence level (t-stat is greater than 2). In other words, when compared to the entire market, the Nicholas Fund, on average, has been slightly smaller and more value oriented. As we mentioned before, being exposed to these types of stocks over longer time horizons has rewarded investors. Once we adjust for these factors, the alpha becomes 0.04% with little to no statistical significance (0.56). Also, the R-Squared measurement gives an indication of how well the regression explains the variation in the returns of the Nicholas Fund. As you can see the 3 Factor Model explains 86% of the variation in the returns of the Nicholas Fund, which is an overwhelming amount. Note that this is based on monthly data.

 

Factor a MKT-TBill SmB HmL  R2
Mean 0.04% 0.91 0.08 0.16 0.86
T-Stat 0.56 50.46 3.11 5.82 -

The chart below gives a visual representation of the Nicholas Fund’s alpha after we adjust for its sensitivities to the small-cap and relative-price premiums. Again, we are looking for not only a positive alpha, but a positive alpha that is statistically significant at the 95% confidence level (green region). As you can see, the Nicholas Fund fails this test for significance. Note that these results are annualized.

Just to further emphasize the point, we are able to replicate the performance of the Nicholas Fund with a blend of IFA Indexes (70% IFA US Large Co. Index/30% IFA US Small Cap Value Index). See the chart below.

As you can see, we gain a very similar return to the Nicholas Fund using a much more cost effective approach, vis a vis, broadly diversified indexes. With no proof of a statistically significant alpha from Nicholas Fund, it would be prudent to pursue the index strategy given its lower cost. 

Conclusion

Whenever the financial services industry is quick to highlight the next best manager, we always approach the story with a healthy dose of skepticism. Why? There is an overwhelming amount of evidence that the vast majority of active managers underperform their respective benchmarks over time. As we have mentioned before, the ability to outperform the market comes down to four things: quicker access to information than most market participants, superior estimation and analytical faculties than the majority of market participants, simply cheating by gaining access to inside information, or random luck. Given the huge growth in the investment management industry filled with some of the best and brightest minds in the world, it is very unlikely that the first two options are viable conclusions. While some hedge fund managers and traders have been prosecuted for acting on inside information, it is probably safe to assume that most “heroes” are just the choices of Lady Luck.