AlphaClone

New Academic Research Supports AlphaClone Investment Approach

New academic research has just been published that further supports AlphaClone's research and investment approach.  The new paper which is entitled "Best Ideas" by Cohen, Polk and Silli concludes that fund managers' high conviction and top holding stock ideas "outperform the market by one to four percentage points a quarter depending on the benchmark used."  Below is the paper’s abstract.

The authors use different definitions for “Best Ideas” but all of them center around the relative proportion a manager allocates to an individual holding and all of them outperform including “popular” Best Ideas (similar to our “Popularity” strategy) and “fresh” Best Ideas (which AlphaClone just calls “Best Ideas”). 

The approach taken above is nearly identical to cloning AlphaClone’s High Concentration fund group – which dynamically selects the 25 funds that have the highest market value spread across the smallest number of holdings (i.e. a high relative portfolio allocation).  We highlighted this group a few weeks ago but its worth repeating that the group does very well across all our clone strategies:

AlphaClone High Concentration Fund Group                                                               (12/31/1999 to 4/17/2009)

Since Inception	Top Holding	Best Idea	Top 10 Pop	S&P 500 TR
# of Holdings	Up to 25	Up to 50	Up to 10	n/a
Annualized Return	8.6%	8.2%	11.4%	-3.7%
Volatility	18.7%	19%	19.6%	16.2%
Sharpe 4%	0.2	0.2	0.4	-0.5
Max Drawdown	-50.4%	-49.7%	-42.4%	-50.9%

 

A couple of weeks ago, we highlighted the fact that being popular in the right circles was potentially very profitable when it comes to clone investing – this group’s Top 10 Popularity clone makes our point again very nicely.  How’s that clone performing YTD?  It’s beating the S&P500 Total Return Index by 23 percentage points.  Yes, 23 percentage points.  Here’s what it holds today:

1              [QCOM]                QUALCOMM INC

2              [V]                       VISA INC

3              [MA]                    MASTERCARD INC

4              [GOOG]                GOOGLE INC

5              [CVS]                   CVS CAREMARK

6              [JPM]                    JP MORGAN CHASE

7              [WMT]                 WAL MART STORES

8              [LMDIA]               LIBERTY MEDIA

9              [MVSN]                MACROVISION

10            [BNI]                    BURLINGTON NORTHERN

------------------- 

ABSTRACT:

"Best Ideas" 

by Cohen, Polk and Silli   

We examine the performance of stocks that represent managers' "Best Ideas." We find that the stock that active managers display the most conviction towards ex-ante, outperforms the market, as well as the other stocks in those managers' portfolios, by approximately one to four percent per quarter depending on the benchmark employed. The results for managers' other high-conviction investments (e.g. top five stocks) are also strong. The other stocks managers hold do not exhibit significant outperformance. This leads us to two conclusions. First, the U.S. stock market does not appear to be efficiently priced, since even the typical active mutual fund manager is able to identify stocks that outperform by economically and statistically large amounts. Second, consistent with the view of Berk and Green (2004), the organization of the money management industry appears to make it optimal for managers to introduce stocks into their portfolio that are not.

Tracking the Hedge Funds with the Lowest Turnover

"Throughout all my years of investing I've found that the big money was never made in the buying or the selling. The big money was made in the waiting."  - Reminiscences of a Stock Operator by Edwin Lefevre

Almost without fail, when we describe AlphaClone to people, one of the first questions we get is about the rate of turnover amongst funds we track and how that affects clone performance. We thought it would be a good idea to address that very issue in this week's article.

First the basics. We track about 250 funds - they are a mix of hedge funds and institutional funds. The table below summarizes turnover statistics for each group:

Turnover Statistics: Hedge Funds vs Institutional Funds*

Fund Type	# of Funds	Turnover  Q4 2008	Turnover Average*
Hedge Funds	148	57.30%	48.90%
Institutional Funds	102	26.80%	20.70%

 *Based on funds tracked by AlphaClone    **Average for prior four quarters

As you can see from the table above, hedge funds tend to turnover twice as much as institutional funds - no surprise there. What is not apparent is that the 48.9% average turnover number for hedge funds in the table above is significantly higher than the historical average for the group, which is about 40%. Hard to predict the future but we expect that turnover for that group will begin to revert to the historical mean as the financial markets stabilize.

In any case, let us get back to how turnover effects clone performance. Turnover figured so prominently in our thinking when we were building AlphaClone that we decided to create a dynamic fund group that selects funds based solely on turnover. The AlphaClone "Low Turnover" fund group selects the 25 funds that have the lowest average turnover over the past four quarters. The group is dynamic so it runs the selection process each quarter (i.e. the funds that make up the group can change each quarter).

Next we looked at how each of our "clone strategies" faired when applied to this specific group. Just to remind you, there are three basic clone strategies:

• Top Holdings: based on the top holdings amongst funds in the group
• Best Ideas: based on the largest new buys amongst the group ("Best Ideas")
• Popularity: based on the stocks that are held by the largest number of funds in the group  

You can vary the number of holdings in each strategy and run backtests for each "variant" (e.g., Top 1 Holding, Top 3 Holdings, exc.) but to keep things simple we've selected one variant from each strategy and summarize the relative performance in the table below:

AlphaClone "Low Turnover" Fund Group (as fo 4/10/2009)

2000-2009	Top 3 Best Ideas	Top 3 Holdings	Top 3 Popularity	S&P 500 Index
Ann. Return (%)	3.6	-0.8	-4.8	-4.3
Volatility (%)	20.0	16.2	21.1	16.0
Sharpe (4%)	0.0	-0.3	-0.4	-0.5
Max Drawdown (%)	-45.2	-49.9	-60.8	-50.9
Best Year (%)	52.7	35.4	33.6	28.7
Worst Year (%)	-34.8	32.9	-36.6	-37.0
Rank*	14 of 22	17 of 22	14 of 25	n/a
# of clone holdings	Up to 50	Up to 75	Up to 3	n/a

* Ranks clone versus other Fund Groups for same clone strategy based on annualized performance since 2000.

Based solely on the results above, it seems that a strategy that follows the "flow" tends to perform well on low turnover funds. The Top 3 Best Ideas clone beat its benchmark by nearly eight percentage points annualized since 2000, with comparable volatility and a lower maximum drawdown.

However, it is also likely that low turnover by itself is probably not the best or only criteria one should use to select funds to clone. All three highlighted strategies for this fund group were in the middle of the pack when ranked versus other fund groups for the same clone straegy based on annualized performance since 2000.

So how can you use the above analysis? Well one way to use it that may prove interesting is to pick out one individual fund from the list of 25 funds that made up the Low Turnover fund group in the most recent quarter and look at the Best 3 Ideas strategy for just that single fund. We did that for Dodge & Cox and its Top 3 Best ideas strategy outperformed the market by 12.3 percentage points annualized since 2000. The bottom line is that in cloning, sometimes "going with the flow" is in fact the "best idea."

Picking Popular Stocks That Return 10%+ Annualized Since 2000

How many times have you heard that “crowded hedge fund trades don’t perform very well”?  Over the past couple of years, research out of Goldman Sachs and others has shown that if you invest in companies that are owned by the largest number of hedge funds each quarter (i.e. most popular holdings) you will tend to underperform the broader market.  Goldman’s basket of “20 most popular” stocks for example tracked the 20 most popular holdings amongst 700+ hedge funds.  It found that over a 22 quarter period (Q3 ‘01 to Q4 ‘06) the basket underperformed the S&P 500 by an average of 17 basis points per quarter.

 While we have no argument with Goldman’s findings, we would suggest that the question isn’t so much whether a stock is popular or not, but, in what group is the stock popular?  To illustrate our point, the table below summarizes the annualized returns since 2000 of seven Top 10 Popularity clones available on AlphaClone.  As a reminder the Top 10 Popularity clone invests in the group’s ten most popular holdings each quarter. 

  Seven Top 10 Popularity Clones                                                         Annualized Return (1/1/2000 to 4/3/2009)

Fund Group	Clone  Return	Benchmark Return	+/-
Tiger Cubs	11.7%	-4.1%	15.8%
Concentrated Funds	10.3%	-4.1%	14.4%
Sector Picks: Basic Materials	11.4%	-1.0%	12.4%
Sector Picks: Consumer Non-Cyclical	13.3%	1.1%	12.2%
Sector Picks: Capital Goods	6.0%	-5.6%	11.6%
Sector Picks: Energy	13.2%	6.3%	6.9%
Hedge Fund Index	-2.9%	-4.1%	1.2%

 

Note that how a clone performs is very much correlated to the criteria used to define the group from which its popularity is then derived.

The Hedge Fund Index group is the poorest performer in the table , beating the S&P 500 Total Return Index by 1.2 percentage points annualized over the analysis period.  Interestingly, just as the Goldman basket was based on a broad universe of 700+ hedge funds, this group selects popular stocks from the broadest data set of any other group in the table because it includes all hedge funds in our database (about 140) and doesn’t employ screens for sector or company size. 

But don’t discount the Hedge Fund Index group just yet.  The four Sector Picks groups in the table are based on the same universe of funds only screened for a specific sector and their popularity clones perform extremely well versus their sector benchmark index (we use S&P 500 Sector Indexes). 

Last but by no means least, being popular amongst our Concentrated Funds group (those with the largest market values spread out across the fewest positions) or our Tiger Cubs group (former Julian Robertson protégés) means you returned 15.8 and 14.4 percentage points respectively in annualized excess returns since 2000 - stellar by any measure. 

If anything, the above analysis goes to prove that being popular in investing is all about who your fan base is which is why we think that being able to select and clone your own fund groups is one of AlphaClone’s most valuable features.

All portfolios in the table above are available to AlphaClone Full Members.  The Tiger Cubs group is also available to Guest Pass Members and Basic Members.